Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 7 Jan 2009 08:42:06 -0800
From:      Joel Jacobson <jjacobson@panasas.com>
To:        Ivan Voras <ivoras@freebsd.org>
Cc:        freebsd-geom@freebsd.org
Subject:   Re: performance problem with gstripe
Message-ID:  <F89B4A24-E877-47E1-89FD-94F1D91761CE@panasas.com>
In-Reply-To: <gk19r9$p9m$1@ger.gmane.org>
References:  <4AD370A6-2226-442F-BD80-8CFD4045B094@panasas.com>	<gk1664$g6d$2@ger.gmane.org> <CA9F4E99-5819-4FDB-B8F4-8AA21759942B@panasas.com> <gk19r9$p9m$1@ger.gmane.org>

next in thread | previous in thread | raw e-mail | index | archive | help
still works badly at 64k, but works well if i use 32k (and have  
kern.geom.stripe.fast=1).  that being said, i was only seeing 64k I/O  
through ufs when i was doing the 256k stripe, so im still not sure why  
this matters.

i have a somewhat hidden agenda here, too, in that i have my own  
filesystem that suffers the same problem im seeing with dd.  i figured  
there was something ufs does which i do not, and was trying to figure  
out what that might be.  it works fine on 4.6.2 using ccd and a 256k  
stripe size [and i send 128k I/O requests, which is what i would  
prefer to see sent to the driver, rather than 64k].

- j

On Jan 6, 2009, at 8:08 PM, Ivan Voras wrote:

> Joel Jacobson wrote:
>> here's what i did:
>
>> ca-sbox-2# dd if=/dev/zero of=/dev/stripe/d0 bs=512k count=4096
>> 4096+0 records in
>> 4096+0 records out
>> 2147483648 bytes transferred in 34.124683 secs (62930508 bytes/sec)
>
>> ca-sbox-2# dd if=/dev/zero of=/mnt/bigfile bs=512k count=4096 &&
>> /usr/bin/time sync
>> 4096+0 records in
>> 4096+0 records out
>> 2147483648 bytes transferred in 11.081184 secs (193795502 bytes/sec)
>
>
> Hmm ok, you might be hitting the MAXPHYS problem. Could you try and
> create your gstripe array with a stripe size of 64 kB and 32 kB and  
> test
> on those ?
>




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?F89B4A24-E877-47E1-89FD-94F1D91761CE>